Amplitude Modeling dbt Package (Docs)

What does this dbt package do?

• Produces modeled tables that leverage Amplitude data from Fivetran's connector in the format described by this ERD and builds off the output of our Amplitude source package.

• Enables users to do the following:

  • Leverage event data that is enhanced with additional event type and pivoted custom property fields for later downstream use
  • View aggregated metrics for each unique session
  • View aggregated metrics for each unique user
  • View daily performance metrics for each event type
  • Use the enhanced event data to leverage dbt metrics to generate additional analytics
  • Incorporate the dbt Product Analytics package to further enhance Amplitude data, like for funnel and retention analysis

This package also generates a comprehensive data dictionary of your source and modeled Amplitude data through the dbt docs site. You can also refer to the table below for a detailed view of all tables materialized within this package by default.

How do I use the dbt package?

Step 1: Prerequisites

To use this dbt package, you must have the following:

• At least one Fivetran Amplitude connector syncing data into your destination.
• A BigQuery, Snowflake, Redshift, PostgreSQL, or Databricks destination.

Databricks dispatch configuration

If you are using a Databricks destination with this package, you must add the following (or a variation of the following) dispatch configuration within your dbt_project.yml file. This is required in order for the package to accurately search for macros within the dbt-labs/spark_utils then the dbt-labs/dbt_utils packages respectively.

dispatch:
  - macro_namespace: dbt_utils
    search_order: ['spark_utils', 'dbt_utils']

Database Incremental Strategies

This package's incremental models are configured to leverage the different incremental strategies for each supported warehouse.

For BigQuery and Databricks All Purpose Cluster runtime destinations, we have chosen insert_overwrite as the default strategy, which benefits from the partitioning capability.

For Databricks SQL Warehouse destinations, models are materialized as tables without support for incremental runs.

For Snowflake, Redshift, and Postgres databases, we have chosen delete+insert as the default strategy.

Regardless of strategy, we recommend that users periodically run a --full-refresh to ensure a high level of data quality.

Step 2: Install the package

Include the following Amplitude package version in your packages.yml file:

TIP: Check dbt Hub for the latest installation instructions, or read the dbt docs for more information on installing packages.

packages:
  - package: fivetran/amplitude
    version: [">=0.5.0", "<0.6.0"] # we recommend using ranges to capture non-breaking changes automatically

Do NOT include the amplitude_source package in this file. The transformation package itself has a dependency on it and will install the source package as well.

Step 3: Define database and schema variables

By default, this package will run using your target database and the amplitude schema. If this is not where your Amplitude data is, add the following configuration to your root dbt_project.yml file:

# dbt_project.yml

...
config-version: 2

vars:
    amplitude_database: your_database_name    
    amplitude_schema: your_schema_name

Step 4: Configure event date range

Because of the typical volume of event data, you may want to limit this package's models to work with a recent date range. However, note that the amplitude__daily_performance, amplitude__event_enhanced, and amplitude__sessions final models are materialized as incremental tables.

The default date range for the stg_amplitude__event model starts at '2020-01-01' and ends one month past the current day. Conversely, for the date spine model in this package the default date range starts at 2020-01-01 and ends one day after the current day. To customize the date range, add the following configurations to your root dbt_project.yml file:

# dbt_project.yml
...
vars:
    amplitude__date_range_start: '2022-01-01' # your start date here
    amplitude__date_range_end: '2022-12-01' # your end date here

If you adjust the date range variables, we recommend running dbt run --full-refresh to ensure no data quality issues within the adjusted date range.

(Optional) Step 5: Additional configurations

vars:
  amplitude:
    lookback_window: number_of_days # default is 3

# dbt_project.yml
...
config-version: 2
vars:
    <package_name>__<default_source_table_name>_identifier: your_table_name

# dbt_project.yml
models:
    amplitude_source:
      +schema: my_new_schema_name # leave blank for just the target_schema

# dbt_project.yml
...
vars:
    event_properties_to_pivot: ['event_property_1','event_property_2']
    group_properties_to_pivot: ['group_property_1','group_property_2']
    user_properties_to_pivot: ['user_property_1','user_property_2']

(Optional) Step 6: Using this package with the dbt Product Analytics package

packages:
  - package: mjirv/dbt_product_analytics
    version: [">=0.1.0"]

-- # product_analytics_funnel.sql
{% set events =
  dbt_product_analytics.event_stream(
    from=ref('amplitude__event_enhanced'),
    event_type_col="event_type",
    user_id_col="amplitude_user_id",
    date_col="event_day",
    start_date="your_start_date",
    end_date="your_end_date")
%}

{% set steps = ["event_type_1", "event_type_2", "event_type_3"] %}

{{ dbt_product_analytics.funnel(steps=steps, event_stream=events) }}

(Optional) Step 7: Orchestrate your models with Fivetran Transformations for dbt Core™

Does this package have dependencies?

This dbt package is dependent on the following dbt packages. These dependencies are installed by default within this package. For more information on the following packages, refer to the dbt hub site.

IMPORTANT: If you have any of these dependent packages in your own packages.yml file, we highly recommend that you remove them from your root packages.yml to avoid package version conflicts.

packages:
    - package: fivetran/amplitude_source
      version: [">=0.3.0", "<0.4.0"]

    - package: fivetran/fivetran_utils
      version: [">=0.4.0", "<0.5.0"]

    - package: dbt-labs/dbt_utils
      version: [">=1.0.0", "<2.0.0"]

    - package: dbt-labs/spark_utils
      version: [">=0.3.0", "<0.4.0"]

How is this package maintained and can I contribute?

Package Maintenance

The Fivetran team maintaining this package only maintains the latest version of the package. We highly recommend you stay consistent with the latest version of the package and refer to the CHANGELOG and release notes for more information on changes across versions.

Opinionated Decisions

In creating this package, which is meant for a wide range of use cases, we had to take opinionated stances on a few different questions we came across during development. We've consolidated significant choices we made in the DECISIONLOG.md, and will continue to update as the package evolves. We are always open to and encourage feedback on these choices, and the package in general.

Contributions

These dbt packages are developed by a small team of analytics engineers at Fivetran. However, the packages are made better by community contributions.

We highly encourage and welcome contributions to this package. Check out this post on the best workflow for contributing to a package.

Are there any resources available?

• If you encounter any questions or want to reach out for help, see the GitHub Issue section to find the right avenue of support for you.
• If you would like to provide feedback to the dbt package team at Fivetran, or would like to request a future dbt package to be developed, then feel free to fill out our Feedback Form.